Issuu

International Research Journal of Engineering and Technology (IRJET)

e-ISSN: 2395-0056

Volume: 12 Issue: 12 | Dec 2025

p-ISSN: 2395-0072

www.irjet.net

Sign Language Translator Using Artificial Intelligence Prof. M. D. Choudhari, Chinmay Pathak, Gunmay Kharabe, Rushikesh Bhise, Shishir Mane Department of Artificial Intelligence & Data Science K. D. K. College of Engineering Nagpur, India Department of Artificial Intelligence & Data Science K. D. K. College of Engineering Nagpur, India -----------------------------------------------------------------------------***--------------------------------------------------------------------

Abstract- A significant communication divide separates users of sign language from those who do not, often leading to the social and professional isolation of the deaf and hard-of-hearing community. This gap is widened by a shortage of real-time, portable, and affordable translation technologies. This project introduces a system engineered to overcome these barriers by translating sign language into text or speech, and vice versa, in real time. The primary aim is to facilitate seamless communication between signers and non-signers. By doing so, the project anticipates fostering greater inclusivity, enhancing independence for deaf individuals, and enabling their full participation in all social and professional environments. Keywords—sign language recognition, computer vision, deep learning, human-computer interaction, continuous SLR, transformers, accessibility technology Introduction

INTRODUCTION

Communication is a fundamental pillar of human society. For the deaf and hard-of-hearing community, sign language— a visual language using hand gestures, facial expressions, and body movements—is a primary means of communication. However, a significant communication gap exists between sign language users and those who do not know sign language. This barrier can lead to social and professional exclusion, limiting opportunities in education, workplaces, and daily life The reliance on human interpreters is not always feasible or accessible, highlighting an urgent need for an affordable, portable, and accurate technological solution. AI-powered Sign Language Recognition (SLR) systems aim to address this need directly. A Live Sign Language Translator captures gestures in real-time using a standard camera, employing computer vision, machine learning, and natural language processing to recognize and translate them. The goal is to create a seamless, two-way communication tool that converts sign language into text or speech, and can also convert speech into sign language representations. Such a system promotes greater accessibility and independence for its users. Beyond accessibility, this project highlights how AI can be applied for social good, promoting inclusivity and equal participation in society. It can be used in schools to support deaf students, in hospitals to assist communication between doctors and patients, and in workplaces to create inclusive environments. Moreover, as wearable and mobile devices evolve, such translators can become portable and user-friendly, making them widely accessible. By eliminating communication barriers, it empowers individuals with hearing impairments to interact freely, gain equal opportunities, and live more independently. This project demonstrates the true potential of AI when applied to human needs, combining technology with compassion for a better society. This paper presents a comprehensive review of the state-of-the-art in AI-based SLR. We provide a structured overview of the technological evolution, methodologies, and persistent challenges in the field. The key contributions of this review are: 1.A detailed survey of modern deep learning architectures for SLR. 2.An analysis of the role of datasets and the critical bottlenecks related to data scarcity. 3.A thorough discussion of the open challenges defined by the core problem statement. 4.An outlook on future research directions required to build a user-friendly and effective system with low latency that requires no special hardware like gloves or sensor.

II.

PROBLEM STATEMENT AND OBJECTIVES

The central problem is the lack of immediate and accessible communication between signers and non-signers. This leads to several interconnected issues: A. Social and Professional Exclusion: Communication barriers are a primary cause of exclusion for the deaf community. B. Interpreter Dependency: The reliance on human interpreters is a significant bottleneck, as they are not always available or affordable. C. Technological Gaps: Existing translation tools are often expensive, not portable, or lack real-time accuracy. D. Linguistic Barriers: Inconsistencies across regional sign languages can lead to miss communication and learning sign language presents time and resource constraints for non-signers.

Impact Factor value: 8.315

ISO 9001:2008 Certified Journal

Page 422